(163d) Model-Based Economic Optimisation of CO2 Compressor Train Design and Operation

Compressor trains are typically currently designed using heuristics, in an iterative fashion. This paper describes an approach for rapidly determining optimal design parameters of a CO₂ compression train for Carbon Capture and Storage (CCS) using rigorous model-based optimisation methods that can take into account time-varying flowrates or delivery pressures. Optimisation variables can include continuous decision variables such as inter-stage pressures, and integer or discrete decision variables, such as number of sections.

Rigorous models of a single-section compressor were constructed in the gPROMS advanced process modelling environment. These were used to construct a superstructure model that contained a maximum-allowable set of inter-connected compressor sections (arbitrarily set to 8 sections), including recycle loops and intercoolers. This was then subjected to a set of optimisations, with an objective function comprising total annualised capital and operating costs. Execution involved solution of a Mixed Integer Non-Linear Programming (MINLP) optimisation problem using the Outer Approximation / Equality Relaxation / Augmented Penalty (OAERAP) algorithm included in gPROMS platform.

The first optimisation determined the optimal number of sections, section diameters, required heat transfer area of the coolers and other key design attributes for a given service, based on identical pressure ratios across each section. The second performed the same optimisation, but with the inclusion of inter-section pressures as optimisation variables. The third case applied a multi-period optimisation approach to determine the optimal number of sections for a typical daily load profile that emulated the typical flow of CO₂ for a power station following grid demand.

The approach makes it possible accurately to determine trade-offs between higher capital expenditure (e.g. the use of more compressor sections) on the one hand, and increased flexibility of service and lower operating costs (through the application of optimised inter-section pressures), on the other hand, to minimise the overall annualised cost of operation under varying flowrate conditions.